Related Experiment Video
Updated: Jun 15, 2026
03:55
Computer-Aided Three-Dimensional Visualization in the Treatment of Locally Advanced Thyroid Cancer
Published on: June 9, 2023
Evaluation of automatic atlas-based lymph node segmentation for head-and-neck cancer.
Liza J Stapleford1, Joshua D Lawson, Charles Perkins
1Department of Radiation Oncology, Emory University School of Medicine and Winship Cancer Institute of Emory University, Atlanta, GA 30322, USA.
Summary
Automatic atlas-based lymph node segmentation (LNS) for head and neck cancer is accurate and efficient. This method reduces inter-observer variability compared to manual segmentation, saving time in treatment planning.
Area of Science:
- Medical imaging and radiation oncology.
- Computational anatomy and image analysis.
Background:
- Accurate lymph node segmentation (LNS) is crucial for effective head and neck cancer radiotherapy.
- Manual segmentation is time-consuming and prone to inter-observer variability.
Purpose of the Study:
- To assess the accuracy, efficiency, and inter-observer variability of automatic atlas-based LNS compared to manual methods.
- To determine if automatic LNS improves treatment planning efficiency.
Main Methods:
- Five physicians manually contoured lymph node volumes on CT scans from 5 patients.
- Automatic contours were generated using an atlas-based approach and subsequently modified by physicians.
- The Simultaneous Truth and Performance Level Estimation (STAPLE) algorithm was used to establish a reference "true" segmentation.
Main Results:
- Automatic contours showed high accuracy, comparable to manual segmentation (e.g., 76% Dice similarity coefficient).
- Automatic-modified contours significantly reduced contour volume range and false positivity compared to manual contours.
- Average time savings of 11.5 minutes per patient (35% reduction) were achieved with automatic segmentation.
Conclusions:
- Atlas-based automatic LNS for head and neck cancer is accurate and efficient.
- This automated approach effectively reduces inter-observer variability in contouring.
- Automatic LNS offers a promising improvement over manual segmentation for treatment planning.
